During the VISIT Satellite Chat on the morning of December 12, forecasters at the WFO in Columbia, SC alerted us to a fog event that took place in their CWA during the early morning hours of December 3, 2012. During the discussion we showed different types of imagery and products that might be useful to help diagnose and predict fog, and below we expand on some of these.
A relatively new product that is currently being used at some WFOs is model forecasts of satellite imagery, made possible by using a high-resolution model combined with a radiative transfer model. One of the motivations for producing this synthetic satellite imagery came from the GOES-R Proving Ground, which seeks to allow forecasters to test and evaluate potential imagery and products that will be available when the GOES-R satellite is launched around late 2015. The idea is that synthetic imagery can be used to show different bands and products that are not available today from the current GOES satellites. Conventional imagery is also generated, in order to allow forecasters to compare predicted imagery with actual satellite imagery. One of the products that is produced emulates the legacy fog product that is currently available on AWIPS. A loop of this synthetic low cloud / fog product between 01:00 and 08:00 UTC December 3 is shown here:
The 4-km NSSL WRF-ARW model is used to generate this synthetic imagery. The model is run once per day (at 0000 UTC) out to 36 hours. The model output is used as input to a radiative transfer model that calculates brightness temperatures at given wavelengths. After the 3.9 um and 10.35 um bands have been created, they are differenced, similar to the GOES legacy fog product. In this color table, low clouds appear blue (positive temperature difference), and high clouds are black (negative temperature difference). This is from the model run on 0000 UTC December 2. Note that this model forecast does well in delineating the low clouds from the high clouds.
We can compare the synthetic imagery with the familiar GOES low cloud / fog product which is simply the difference between the 10.7 um and 3.9 um bands:
Note the similarities and differences across South Carolina with respect to the low cloud / fog (blue) location and timing.
One important thing to remember when looking at the synthetic low cloud / fog product is that it does NOT discriminate between low clouds and fog. We overlaid the ceiling and visibility observations onto the GOES fog product loop to help address this question.
Another way to address this question is to look at the cloud cover field at the lowest vertical level in the model, this would indicate fog forecast by the model:
Note the model forecast of fog development across South Carolina during this time period.
Real-time imagery may be found on the following web-sites:
Synthetic low cloud / fog from the NSSL 4-km WRF-ARW:
Cloud cover at the lowest vertical level from the NSSL 4-km WRF-ARW:
While synthetic satellite imagery is not routinely produced at this time by most models, there are other forecast products available from high-resolution models that attempt to address fog occurrence. An example is shown below for this case from the HRRR (High Resolution Rapid Refresh) model, available online at http://ruc.noaa.gov/hrrr/
Displayed is a 4-hour forecast of surface visibility from the 0400 UTC HRRR run on 3 December, valid at 0800 UTC. A large area of visibility below 0.5 miles is predicted, including over much of South Carolina. Other applicable products (not shown) include cloud base height and aviation flight rules.